Owning copyright and proving you created the work are separate problems, and UK law solves only the first: protection arises automatically at creation, while the evidence needed to enforce it is left entirely to the creator. A claim can fail on that second problem even when the right itself is beyond question.
A freelance copywriter in Manchester spends three weeks producing a campaign concept for a prospective client. The pitch is polished, the wording precise, the strategy distinct. The client says they need time to decide. Six weeks later, the copywriter sees the concept — almost word for word — running as an in-house campaign, with no commission, no credit, and no acknowledgement of its source.
A self-published novelist in Bristol uploads chapter drafts to a writing community throughout 2024. Two years later, a traditionally published debut arrives carrying characters and phrases that closely mirror her unpublished work. The published novel's release date is firmly documented. The dates on her own drafts, sitting on her laptop, are not documented anywhere she does not control.
A Sheffield software engineer commits a novel algorithm to a private repository in March. In September, a competitor product launches with what appears to be the same approach. The engineer is certain of her dates. Nobody else can vouch for a single one of those dates.
In all three cases, the question is not whether the creator owns the work — they do. The question is what happens when ownership meets an opponent with a documented date and a legal team. This guide covers what an authorship dispute demands in evidence, why the methods most creators rely on collapse under challenge, and what a record that survives challenge requires.
The Copyright, Designs and Patents Act 1988 grants copyright the instant a work is made — no registration, no fee, no formality of any kind. Ownership follows authorship under section 11: the person who creates the work owns it, unless they created it as an employee in the course of employment. That default matters more than most freelancers realise — a commissioning client does not own the copyright in commissioned work unless the contract expressly assigns the copyright. The Manchester copywriter owns that campaign concept outright. The client who ran it without a commission was not exploiting a grey area; on these facts, they were infringing.
None of that self-executing generosity extends to evidence. The Act creates the right in silence, leaves no public record of the work, its date, or its author, and then places the burden of proving all three on the person asserting the right, to the civil standard of the balance of probabilities. The result is a right that costs nothing to acquire and can cost everything to demonstrate.
The gap is widest where formal publishing infrastructure is most absent — and the creative sector is overwhelmingly composed of the kinds of workers who never encounter that infrastructure. According to DCMS economic estimates, self-employment accounted for 28 per cent of creative industries employment in the year to March 2025, double the 14 per cent rate across the UK workforce as a whole. In March 2025, 93 per cent of UK creative businesses were microbusinesses. A traditionally published novelist has ISBNs, deposit copies, a publisher's records, and a paper trail built into the trade. The copywriter, the photographer, the self-releasing musician, and the independent software developer have none of those systems — and the 1988 Act supplies nothing in their place.
Two further traps compound the problem for collaborative and freelance work. Where two or more creators contribute without their contributions being distinct, the Act makes them joint authors — and years later, a loose collaboration with no records becomes an ownership dispute inside the authorship dispute. And even the right to be credited is not self-executing: the Act requires the attribution right to be asserted, in writing, before an infringer is bound. The pattern is consistent throughout the statute. The rights arrive automatically; every mechanism for evidencing or activating them demands a deliberate act the creator must think to perform.
An original-authorship dispute turns on three questions, and the claimant needs independent evidence for each. First, what is the work? Not a description or a summary — the exact creative material, in the form it took at creation. Second, when did it exist? The claimant's version must demonstrably pre-date the alleged copy, and a court will not take the claimant's word for the date. Third, who was the author? Authorship, or a lawful chain of assignment from the author, must be established.
The three questions interlock. English courts approach copying through inference: where the claimant shows substantial similarity between the works and shows that the defendant had access to the original, the evidential burden shifts to the defendant to explain the resemblance — the approach confirmed by the House of Lords in Designers Guild v Russell Williams. What must have been taken is a substantial part of the original, judged by the quality and originality of what was copied rather than the quantity — a handful of distinctive sentences can qualify where pages of the generic would not. Every element of that inference depends on priority. Similarity proves nothing unless the claimant's work came first, and access cannot be shown to a work that cannot be dated. A claimant who cannot fix their work in time never gets the inference started.
The Bristol novelist's problem lives exactly here. The resemblance between the two novels may be striking. But the published book carries an unimpeachable date, and her drafts carry dates only her own laptop asserts. Unless she can independently establish that her chapters existed first, the inference that should favour her runs in reverse: the documented work looks prior, and hers looks like the echo.
Five kinds of evidence appear again and again when creators try to prove priority, and each carries a structural weakness an opposing legal team will find.
Five common methods — and the structural weakness in each
Method | The structural weakness |
Posting the work to yourself (the 'poor man's copyright') | Envelopes can be opened and resealed, and UK copyright law gives the method no special evidential status |
Email timestamps | Metadata is editable by anyone with server access, and headers can be forged; courts treat the weight as low |
File dates on your own devices | The date reflects one machine and moves with system settings, re-saves, and transfers |
Witness testimony | Tested under cross-examination, and recollection of dates from years past is notoriously unreliable |
Internal drafts and version histories | Produced and held by the claimant, so a court cannot treat the timing as independent |
None of these is worthless; each adds contextual weight, and version histories in particular can corroborate a wider account. But every one of them shares the same flaw the courts keep encountering: the record either lives under the claimant's control or cannot fix content and date together. What a dispute demands is a record made when the work existed, held beyond the claimant's reach, that binds the exact content to the exact moment — and none of the five delivers all three at once.
Watch how the weakness plays out under challenge. The claimant produces a file dated March; the opposing side asks who set the clock on that machine, whether the file has been re-saved since, which backup it was restored from, and why the metadata shows a later modification. None of those questions needs an answer to do damage — each one converts a date into a debate. Evidence that needs its owner's testimony to hold together is evidence the owner must defend line by line.
Strip back any form of evidence that fares well in authorship disputes and the same three characteristics appear. Somebody other than the claimant made the record. The record came into existence alongside the work rather than being assembled after the dispute began. And the record can be tested — a tribunal, or the other side's experts, can check it without trusting anyone's account.
Structured systems delivering all three have existed for decades, but only for narrow categories of work. Publishing infrastructure delivers it for books, through ISBNs, deposit copies, and the paper trail of the trade. Patent filing delivers it for inventions, at the price of cost, delay, and full public disclosure. Notarisation delivers it for individual documents, at a per-document fee and an in-person process that collapses at any real volume. Nothing in that list was built for the everyday output of a working creator — the chapter drafts, the campaign iterations, the design explorations, the commit history, the half-finished stems. Cryptographic timestamping was built for precisely that gap.
Cryptographic prior proof at the qualified standard works in four layers:
Note what the fingerprint approach also protects: privacy. The hash reveals nothing about the file's contents and cannot be reversed into them, so the record fixes the work without disclosing the work — the unpublished manuscript stays unpublished, the unreleased code stays private, and the proof exists anyway.
Run the three scenarios again with such records in place. The copywriter's pitch was hashed and timestamped before the file was ever sent, so the concept's content and date stand on a certificate the client never touched. The novelist's chapter drafts carry independent dates through 2024, and the inference of copying now runs the way the resemblance suggests: her work first, the debut second. The engineer's algorithm carries proof of existence from March that does not depend on any employer system, any platform's retention policy, or her own machine surviving to trial.
For most working creators, the realistic arena for a dispute is not the High Court. Copyright claims up to £10,000 can be brought in the small claims track of the Intellectual Property Enterprise Court, a forum designed for parties without legal teams, where costs recovery is tightly capped and a self-contained evidence bundle does most of the work a barrister otherwise would. In that setting the bundle carries the case: the work as deposited, the certificate binding its content to its date, and a short timeline connecting both to the alleged infringement. The difference between asserting a date and enclosing independent proof of one is frequently the difference between a claim worth bringing and a claim worth abandoning.
One honest limitation belongs in any working guide. A record of priority proves that the work existed, in that form, on that date, in the depositor's hands — no more. It does not prove the work was original, and it does not decide whether what a rival took amounts to a substantial part; those remain arguments for the merits. What the record does is take the fight over dates off the table entirely, so the dispute can be about the only thing worth disputing.
The larger effect arrives before any tribunal sees anything. Most authorship disputes end in correspondence, and the letter that encloses an independently verifiable record of what existed and when reads very differently from a letter of bare assertion. Strong evidence does not just win cases; it shortens disputes, and often prevents them entirely.
The range of parties asking the priority question is also widening. A creator's opponent was once a rival creator or a former client; increasingly the question arrives from AI dataset disputes, licensing negotiations, and due-diligence exercises, where the ability to demonstrate ownership of a body of work — quickly, at scale, without litigation — determines whether the work can be licensed, sold, or defended at all. The evidence that wins a plagiarism claim and the evidence that satisfies a licensing marketplace turn out to be the same evidence.
The habit this asks of a creator is small, and it works best at fixed moments rather than as an afterthought: the first complete draft, the version that leaves your hands for anyone else's, and each revision substantial enough that losing its date would hurt. A pitch should carry a record from before the meeting; a manuscript from before the first submission; a codebase from before the first demo. Collaborative projects benefit twice over: a deposit made before material is shared fixes what each contributor brought and when, closing off the joint-authorship ambiguity before it can open. Treat the record as one more step in finishing the work, alongside the save and the backup. Copyright arrives with the work, automatically and invisibly. Proof arrives only on purpose.
UK Intellectual Property Office: Copyright — official guidance
Copyright, Designs and Patents Act 1988 — full text
CPR 32.19: What UK Courts Require to Prove a Document Is Authentic
This post provides general information about the role of cryptographic evidence. It is not legal advice. For advice on a specific matter, consult a qualified lawyer in your jurisdiction.
James Snell is the founder of Provlyn, a platform providing cryptographic prior proof of IP ownership. provlyn.com